ThreadIndicies constructor was templated on the invocation type, which created
thousand's of versions of that symbol which all had the same behavior. So now
remove that and move that logic into a Worklet function since it requires
the invocation info.
Scatter in worklets
Add the functionality to perform a scatter operation from input to output in a worklet invocation. This allows you to, for example, specify a variable amount of outputs generated for each input.
See merge request !221
This has been requested on the mailing list to make it easier to
interpolate integer vectors.
There are a couple of downsides to this addition. First, it implicitly
casts doubles back to whatever the vector type is, which can cause a
loss of precision. Second, it makes it more likely to get overload
errors when multiplying with Vec. In particular, the operator to cast
Vec of size 1 to the component class had to be removed.
The tetrahedralize algorithms have been changed to use the Scatter
classes to build indices rather than build them on their own.
To implement this efficiently with structured grids, a new ScatterUniform
class was made. I also added a new execution argument tag that allows
you to get the thread indices object from within the worklet.
Previously, each VecFromPortalPermute (the type that held the from field
values) held its own copy of the indices. For point to cell on
structured grids, this was a lot of repeated data values, which has the
potential to fill up cache and registers. Instead, just use pointer
references.
This changes the interface to the ThreadIndices classes to have both
input and output indices. It also adds a visit index to ThreadIndices.
Also added the VisitIndex execution signature tag, which relies on this
behavior.
This now allows for even more efficient construction of uniform point
coordinates when running under the 3d scheduler, since we don't need to go
from 3d index to flat index to 3d index, instead we stay in 3d index
Previously, all Fetch objects received an Invocation object in their
Load and Store methods. The point of this was that it allowed the Fetch
to get data from any of the execution objects. However, every Fetch
either just got data directly from its associated execution object or
else used a secondary execution object (the input domain) to get indices
into their own execution object.
This left two potential areas for improvement. First, pulling data out
of the Invocation object was unnecessarily complicated. It would be much
nicer to get data directly from the associated execution object. Second,
when getting index information from the input domain, it was often the
case that extra computations were necessary (particularly on structured
cell sets). There was no way to share the index information among
Fetches, and therefore the computations were replicated.
This change removes the Invocation from the Fetch Load and Store.
Instead, it passes the associated execution object and a new object type
called the ThreadIndices. The ThreadIndices are customized for the input
domain and therefore have all the information needed for a redirected
lookup. It is also a thread-local object so it can cache computed
indices and save on computation time.
This is to be used in place of BOOST_STATIC_ASSERT so that we can
control its implementation.
The implementation is designed to fix the issue where the latest XCode
clang compiler gives a warning about a unused typedefs when the boost
static assert is used within a function. (This warning also happens when
using the C++11 static_assert keyword.) You can suppress this warning
with _Pragma commands, but _Pragma commands inside a block is not
supported in GCC. The implementation of VTKM_STATIC_ASSERT handles all
current cases.
The boost assert macros seem to have an issue where they define an
unused typedef. This is causing the XCode 7 compiler to issue a warning.
Since the offending code is in a macro, the warning is identified with
the VTK-m header even though the code is in boost. To get around this,
wrap all uses of the boost assert that is causing the warning in the
third party pre/post macros to disable the warning.
In the special case where you are loading the point coordinates for a
structured grid in a point to cell map (an important use case), create a
VecRectilinearPointCoordinates rather than build a Vec of the values.
This will activate the cell specalizations in previous commits.
These changes also added some flat-to-logical index conversion and vice
versa in ConnectivityStructuredInternals. This change also fixed a bug
in getting cells attached to points in 2D grids. (Actually, technically
someone else fixed it and checked it in first. The changes were merged
during a rebase.)
I also added a specalization to Vec for 1D that implicitly converts
between the 1D Vec and the component. This can be convenient when
templating on the Vec length.
Previously, when you requested a CellShape in the ExecutionSignature,
you just got an ID stored in a vtkm::IdComponent. This change returns a
cell shape tag of the appropriate type (or generic if the type is not
known at compile time). This will allow functions called from the
worklet to specialize on the cell type better.
This class was used to store a group of from field data in a topology
map. However, the fetching has been changed to use a customized class
for each type of fetch that can be optimized for the fetch type and does
not require to know the number of items in the fetch at compile time.
Thus, this class is no longer needed, so it is being removed.
This change removes the requirement to specify some maximum cell length
in each of the worklets, which is basically impossible. It also makes
some of the loading more lazy, which might help reduce the number of
registers required in a worklet.
We want to be able to get topological connections where it is difficult
to know how many values you get each time. In this change, the type of
the vector holding the from indices is determined from the connectivity
object, and the worklet does not know the type (it must be templated).
Although you do not need to specify the max number for this value set
(you still currently do for field values), we still need to change the
type for explicit sets that uses something that does not rely on the Vec
class. The cell-to-point method also needs a Vec wrapper that allows it
to shorten the vector dynamically.
Previously, all arrays passed to worklets were designated as either
input or output. No in-place operation was permitted. This change adds
the FieldInOut tag for ControlSignature in both WorkletMapField and
WorkletMapTopology that allows you to read and write from the same
array.
In the CellSet and related classes, a connection was referred to by a
"from" topology element and a "to" topology element. However, in the
worklet control signature tags the elements were referred to by "src"
and "dest." To make things consistent, use "from" and "to" everywhere.
Most of VTK-m follows the convention of calling the 0D topology elements
"points" (which follows the convention of VTK). However, there were
several places where they were referred to as "nodes." Make things
consistent by calling them points everywhere.
Also merged some redundant ExecutionSignature tags.
Also moved from vtkm namespace to vtkm::internal namespace. This change
is to then move the structured connectivity classes to the cont and exec
namespaces.
C and C++ has a funny feature where operations on small integers (char
and short) actually promote the result to a 32 bit integer. Most often
in our code the result is pushed back to the same type, and picky compilers
can then give a warning about an implicit type conversion (that we
inevitably don't care about). Here are a lot of changes to suppress
the warnings.
On one of my compile platforms, GCC was giving conversion warnings from
any boost include that was not wrapped in pragmas to disable conversion
warnings. To make things easier and more robust, I created a pair of
macros, VTKM_BOOST_PRE_INCLUDE and VTKM_BOOST_POST_INCLUDE, that should
be wrapped around any #include of a boost header file.
You can use function level statics, but instead you must use class level
statics, this is due to how nvcc treats method statics as being shared
across all threads in a warp.